Bacterial vaginosis (BV) is the most prevalent gynecologic syndrome among women of reproductive age and has been associated with adverse pregnancy and reproductive health outcomes, including low birthweight, preterm birth, pelvic inflammatory disease and increased risk of HIV and other sexually transmitted infections (STIs). In keeping with CDC Health Protection Goal of Healthy People in Every Stage of Life, the long-term goals of this research proposal are to further existing knowledge on the natural history and etiology of bacterial vaginosis in order to promote healthy pregnancy and birth outcomes (Start Strong), improve reproductive and sexual health, and prevent infectious diseases and their consequences among adults (Live a Healthy, Productive, and Satisfying Life). The specific aims of the proposed research are to provide comprehensive epidemiologic information on patterns of BV, determine potentially modifiable risk factors and identify sexual and non-sexual modes of transmission in rural Ugandan women. To address these issues, two complimentary longitudinal studies of vaginal flora changes were conducted in Rakai, Uganda. The first study followed women ages 13-39 for two years and obtained self-collected vaginal swabs at weekly intervals to assess changes in vaginal flora over time. The second study followed polygamous household units (husbands, wives, and other co-resident women aged 13- 49 in the household) at 6 month intervals to assess clustering of BV among co-resident women with a common sex partner (husband) compared to other co-resident women without a common sex partner. Different analytic strategies will be employed to assess changes in vaginal flora and risk factors associated with these changes. Nonparametric regression techniques, summary measures, and transition matrices will be used to describe vaginal flora patterns within individuals over time. Longitudinal regression techniques, such as random-effects models and transition models, that appropriately account for the correlation within women over time and within households will be used to identify risk factors associated with changes in vaginal flora. Finally, random-intercept models will be used to assess (intra-class) correlation of vaginal flora status between women within polygamous households and to determine whether sharing the same sex partner (polygamous husband) or sharing of other (non-sexual) household items (bathing items, towels, etc.) accounts for this correlation. [unreadable] [unreadable] [unreadable]